Generalized Theory and Analysis of Scalar Modulation Techniques for a m x n Matrix Converter

Abstract

The basic requirement of all the scalar modulation strategies employed for matrix converter, is to define the desired output voltages. The output phase voltages, if not optimized, results in underutilization of semiconductor device ratings. For better utilization, the output reference phase voltages are modified such that the sinusoidal nature of line voltages is not altered. Multiphase drives are in consideration and much work is being done and reported on multiphase matrix converter-based system. New modulation strategies for such systems are being formulated. Most of these new strategies are based on scalar approach; which in contrast to the space vector pulse width modulation, require addition of friendly harmonics in order to achieve optimized output reference phase voltages and maximum utilization of the matrix converter semiconductor switches. On the other hand, in space vector pulse width modulation, the maximum voltages are inherently achieved. This paper presents a generalized theory, explained in mathematical as well as graphical manner, defining the optimized output reference phase voltages for any number of input and output phases, whether odd or even, for m × n matrix converter. Further, this paper also explores the Venturini method for m × n case. It is found that the method is extendable to 3 × n matrix converters only. Finally, a simple and generalized algorithm, applicable to all m × n matrix converter configurations, is discussed. Results for 3 × 5 matrix converter are validated by simulation and experimental results